The dual-core phones benchmarked, LG Optimus 3D takes the top

The dual-core phones benchmarked, LG Optimus 3D takes the top
When we wrote our article about the first generation dual-core chipsets way back last summer, they were all still sampling with manufacturers, but now that three of them have gone into devices, we have the chance to observe how they actually fare.

At the time the Texas Instruments OMAP4 4430 chipset seemed the most promising to us, because of its added functionalities for image processing, and optimized battery consumption. We won't mention the dual-core Snapdragon here, as we are yet to handle it in a phone. OMAP4 seemed to offer more than NVIDIA Tegra 2, while not much was known about Samsung's Exynos chipset, which, until recently, was called Orion.

We only know that Samsung switched from the PowerVR graphics chipset line, which was present in its Hummingbird chipset in the Galaxy S, and in the iPhones, to ARM's own quad-core Mali GPU. Little was known exactly how many triangles per second is Mali supposed to push out. Samsung claimed it will give 5x the performance of its previous GPU used, which would have put it at insane speeds. On ARM's website the same 5x claim is made, but in comparison to the old Mali GPU, so go figure.

Of course, performance depends on many variables - at what level are the GPU and memory clocked, the driver's version, software, etc. The Power VR SGX540 in the TI OMAP4 chipset, as found in the LG Optimus 3D, is clocked around 300MHz, 50% higher than the same GPU in the Galaxy S. NVIDIA uses its own eight-core GeForce GPU in Tegra 2, which performs slightly above the Hummingbird GPU, clocked as it is in the Galaxy S.

Enough with the theory, let's see how the LG Optimus 3D with Texas Instruments OMAP4 4430, the Motorola ATRIX 4G and LG Optimus 2X with NVIDIA Tegra 2, and the Samsung Galaxy S II with Exynos performed in reality. Anandtech has managed to run a couple of quick tests for graphics and browser performance, and the results confirm what our own tests showed the other day.

While in the browser benchmarks they are somewhat close, in graphics the OMAP4 chipset eked out about 5-6 frames more on the GLBenchmark 2.0 Pro test, trailed by the Tegra 2 phone, and, surprisingly, by the Samsung Galaxy S II. All tests are relative - in the browser department, for instance, our own Sunspider JavaScript benchmark with the LG Optimus 2X showed 3096ms, which would have put it on top of Anandtech's tests. Nevertheless, in graphics the LG Optimus 3D definitely has the upper hand, but not by much, and we guess the dual-channel LPDDR2 memory controller helps every bit.

Our Quadrant benchmarks from the other day placed LG's finest at the top in general benchmarks too, but we thought it is due to the file system the Samsung Galaxy S II might be using, which slows its read/write performance down. The graphics results, though, are rather revealing, despite that we sincerely doubt these tests have been optimised for two CPU cores, let alone multicore GPUs - Tegra 2 has eight-core GPU, and Exynos has quad-core, for example.

Anyway, these benchmarks are just for bragging rights - any of these handsets has more power than you will ever need in your phone for running the everyday apps, and it will be so for a while.

Besides, although Tegra 2 beat Exynos in the graphic test, it still does Full HD 1080p video at only 24fps, whereas the Galaxy S II does it with 30fps, so synthetic benchmarks are just for kicks. Also, the Optimus 3D and Galaxy S II are non-final units, so take these with a grain of salt. The Motorola ATRIX 4G's software is probably not finalized too, otherwise the company wouldn't have warned that Full HD 1080p recording will come post-launch via a firmware update.

source: Anandtech



1. Kakarotto

Posts: 255; Member since: Dec 07, 2010

Everyone was thinking about Tegra 2 and Orion/Exynos - what will be better. And now we have unexpected winner - TI OMAP 4430.

2. calamazoo unregistered

I don't think we should call 4-5 frames per second more a definite winner, and in the browser mark they also got close results. in comparison to single-core chipsets, though, which get results in the low 30k on this browser test, they are all definitely winners :-)

8. giantgnome

Posts: 67; Member since: Feb 11, 2011

None of these tests have been optimized for the new phones. Once the tests catch up to the hardware the results can change drastically, not to mention when the OS bugs are finally worked out. In any case, they are so close in score I would take the S2 over the Optimus 3D just for two things alone, the S2 is much thinner, and it can do 1080P at 30fps. I don't need the 3D gimmicks.

11. m16 unregistered

Design and sleekness is one aspect but for me and most gadget geeks performance has to be put into more consideration.

3. OG Droid unregistered

Nothing like benchmarking new phones to make my Moto Droid look outdated. Amazing what can happen in 18 months. This just makes my update choices more confusing. Have to wait till I can get my hands on some of these before making a final decision on upgrading.

4. ribbonsalmark

Posts: 84; Member since: Jan 06, 2011

Not as single SE phone on there.

5. snowgator

Posts: 3621; Member since: Jan 19, 2011

Wave of the future I know, but as far as new phone purchases go- how long before this type of processing speed is actually needed for the more basic user? Is it truly a worthwhile for the premium price, say in the next 12 months? Most of what I have read is dual core may be 2 years ahead of the need- is this accurate?

6. Jeromeo

Posts: 135; Member since: Jan 11, 2010

Who would've known that the myTouch could beat out so many 'better' devices with faster processors?! Would have been nice to see the Vibrant (the only American GalaxyS variant with 2.2 froyo) in the lineup. I object to showing a N900 at a measly 600mHz... Mine is clocked at 1150mHz!

7. etretti unregistered

i think Atrix would have same scores or better then lg x2 with the same resolution, kinda unfair bench but sure if u play something its going to be played in the native resolution so...bleh :)

9. susy unregistered

lun te chro .

10. susy unregistered

lun te chro

12. jeffplusplus unregistered

1080P is 24fps, 1080I is 30fps. probably should know what you are talking about before you type.

13. stewy55 unregistered

1080p is a resolution, not a screen refresh rate. a standard lcd tv has a 60hz refresh at 1080p, what would you call that then if p is only 24 fps? some new tvs refresh at 480hz now. 1080p and 1080i are only different in the way the image is processed and displayed on your tv. like a panaramic vs 2 side by side photos. This is done only because a phone or small camera dont have enough image processing power to store the data fast enough.That is the only difference in the 2. maybe you should read your own words Jeffplusplus, and do as instructed.

Latest Stories

This copy is for your personal, non-commercial use only. You can order presentation-ready copies for distribution to your colleagues, clients or customers at or use the Reprints & Permissions tool that appears at the bottom of each web page. Visit for samples and additional information.